[SOLVED] Cant get 6970 to mine? (Locks up)

[daemonic]

Hiya,

Ive been reading and lurking for a while now, and have finally got some bits together to set up a test rig.

My card of choice for the time being is a Sapphire 6970, (6990's arent available anywhere ive tried).

I have a weird problem that occurs for me in both Ubuntu and Centos installs (using the guides ive found on here).

If i setup poclbm to mine using the cpu, it works fine and chugs along at a measly 1200 khash/s, but if i then change the device to use the 6970 (detected as Cayman by poclbm) it works for a second or so, and then seems to lock up. The machine still reponds on command line in another session window, but trying to Ctrl-C the poclbm process fails to stop it unlike in cpu mode when it exits with "bye".

The drivers and sdk are all installed ok as shown by aticonfig;

Code:

[root@localhost ~]# aticonfig --odgc --adapter=all

Adapter 0 - AMD Radeon HD 6900 Series
                            Core (MHz)    Memory (MHz)
           Current Clocks :    250           150
             Current Peak :    880           1375
  Configurable Peak Range : [500-950]     [1375-1450]
                 GPU load :    0%
[root@localhost ~]# aticonfig --odgt --adapter=all

Adapter 0 - AMD Radeon HD 6900 Series
            Sensor 0: Temperature - 42.50 C
[root@localhost ~]# aticonfig --pplib-cmd "get fanspeed 0"
Fan speed query:
Query Index: 0, Speed in percent
Result: Fan Speed: 24%

When running poclbm, i get the following output before the process locks up and whilst the process is locked up i get the following weird output from aticonfig;

Code:

[root@localhost poclbm]# python2.7 poclbm.py -d0 --host=mining.eligius.st --port=8337 --user=[redacted] --pass=x --verbose -v -w128
26/06/2011 01:49:12, Setting pool [redacted] @ mining.eligius.st:8337
26/06/2011 01:49:26, 55 khash/s
26/06/2011 01:49:26, Using new LP URL /LP
26/06/2011 01:49:26, LP connected to mining.eligius.st:8337
26/06/2011 01:49:27, 101985 khash/s

Code:

[root@localhost ~]# aticonfig --odgc --adapter=all

Adapter 0 - AMD Radeon HD 6900 Series
                            Core (MHz)    Memory (MHz)
           Current Clocks :    0           0
             Current Peak :    880           1375
  Configurable Peak Range : [500-950]     [1375-1450]
                 GPU load :    200%
[root@localhost ~]# aticonfig --odgt --adapter=all

Adapter 0 - AMD Radeon HD 6900 Series
            Sensor 0: Temperature - 4294839.50 C
[root@localhost ~]# aticonfig --pplib-cmd "get fanspeed 0"
Fan speed query:
Query Index: 0, Speed in percent
Result: Fan Speed: 100%

As you can see, the temperature is very wrong for starters, can anyone give any pointers as to what could be wrong?

[1714883396]

1714883396

[1714883396]

1714883396

[1714883396]

1714883396

[daemonic]

Just reinstalled the machine to windows 7 pro 64bit, and also installed the drivers and steam sdk.

Ive tried guiminer and this shows a few hundred khash/s and then locks up the whole of windows, surely this isnt right?

Does anyone have any ideas?

[x0Jakeyboy0x]

I highly doubt this is the problem, but have you tried to run the programs with "Run as Administrator"? And does the SDK automatically have OpenCL support? Could you perhaps try a different setting? Cuda, Phoenix, etc in guiminer aswell?

[bcforum]

I used:

http://foreverrising.wordpress.com/2011/04/16/bitcoin-mining-and-ubuntu-10-10-ati-radeon-5xxx/

and it worked well for me on two machines with dual 6970s. You need to get the 2.4 version of the SDK instead of the 2.1 version, but it is unchanged otherwise.

How big is your power supply? I have an 800W for two cards which causes one card to lockup occasionly. My seconds system with and 850W runs without any problems. I swapped cards between slots and machines and with a spare and the same slot always locks up.

[martona]

Just reinstalled the machine to windows 7 pro 64bit, and also installed the drivers and steam sdk.

Ive tried guiminer and this shows a few hundred khash/s and then locks up the whole of windows, surely this isnt right?

Does anyone have any ideas?

It looks like a hardware problem. You could try underclocking the video card to see if you can limp along that way. Have you tried using the cards "normally"? You know, in 3D mode? Run a benchmark or two? If it is indeed a hardware issue it's still not certain that it's the card, it could be a motherboard or PSU problem as well.

[daemonic]

Thanks for the pointers guys/gals, there was enough info between you all for me to go buy a new psu, and hey presto all is now good